Linear beam electron devices are used in sophisticated communication and radar systems that require amplification of a radio frequency (RF) or microwave electromagnetic signal. A conventional klystron is an example of a linear beam electron device used as a microwave amplifier. In a klystron, an electron beam is formed by applying a voltage potential between a cathode emitting electrons and an anode accelerating these emitted electrons such that the cathode is at a more negative voltage with respect to the anode. The electrons originating at the cathode of an electron gun are thereafter caused to propagate through a drift tube, also called a beam tunnel, comprising an equipotential surface, thereby eliminating the accelerating force of the applied DC voltage. The drift tube includes a number of gaps that define resonant cavities of the klystron. The electron beam is velocity modulated by an RF input signal introduced into the first resonant cavity. The velocity modulation of the electron beam results in electron bunching due to electrons that have had their velocity increased gradually overtaking those that have been slowed. Velocity modulation in the gain section of the tube leads to bunching, i.e. the transformation of the electron beam from continuously flowing charges to discrete clumps of charges moving at the velocity imparted by the beam voltage. The beam bunches arrive at the bunching cavity, sometimes called the penultimate cavity, where they induce a fairly high RF potential. This potential acts back on the beam, and serves to tighten the bunch. When the bunches arrive at the output cavity they encounter an even higher rf potential, comparable to the beam voltage, which decelerates them and causes them to give up their kinetic energy. This is converted to electromagnetic energy and is conducted to a load. The tighter the bunching, the higher the efficiency. However, a high degree of space charge concentration interferes with the bunching process and the efficiency. Other things being equal, the higher the perveance of a klystron, the lower the efficiency.
The effect of perveance on the gain of a klystron is different. Although the gain is affected by space charge, it is a stronger function of the total current, which is proportional to the perveance. This suggest that if a beam cross-section were made larger, so that the current density and space charge are reduced, both gain and efficiency would benefit. However, such is not the case because a large beam requires a large drift tube, and the electric fields which couple the beam to the circuit fall off across the beam, leading to poor coupling and a drop in both gain and efficiency. A small beam is therefore necessary, but if the power output required is high, the voltage, rather than the current in the beam must be increased for reasonable efficiency.
Bandwidth is inversely proportional to the loaded Qs of the klystron cavities. In the gain section of the tube, where cavities are stagger-tuned, the cavity Qs are loaded by the beam. The higher the current, the higher the loading, and consequently the lower the Q. It does not matter if a single beam or several beams are traversing the cavity. The output cavity, in particular, must by itself have a bandwidth at least equal to the desired bandwidth of the klystron. For the output cavity to produce good efficiency, this bandwidth becomes proportional to the beam conductance. However this leads to higher perveances, and hence lower efficiency. Consequently, in a single beam klystron the efficiency/bandwidth product is approximately constant.
Given the preceding relationships, the advantage of the multiple beam klystron provides is clear. The current is divided into several beams, each with a low space charge, so that it can be bunched tightly in a small drift tube with good coupling coefficient, and hence high efficiency. The gain-bandwidth product is not constant, but increases with the addition of beams. For the same power and gain, the multiple beam klystron is shorter than a conventional klystron.
Despite the potential advantages of multiple beam klystrons, such devices have only been adapted for certain low power or low frequency applications in which a convergent electron beam is not necessary. In these nonconvergent devices, electron beam focusing is provided by immersing the electron gun and drift tubes in a strong magnetic field which guides the electrons along the magnetic flux lines to the drift tubes. In a nonconvergent electron gun, the diameter of the emitting surface is the same as the electron beam that propagates through the RF device. The nonconvergent electron beams of this class of device have limited current density, which prevent them from developing more power at higher frequencies. The amount of current that can be emitted from the cathode is dependent on the size of the emitting surface and the maximum electron emission density that can be provided by the surface. Maximum electron emission densities from typical cathodes operating in the space charge limited regime are on the order of 10-20 amps/cm^2.
In a convergent electron gun, the cathode diameter exceeds the diameter of the final electron beam, which means that more current can be provided. The current gain is proportional to the area compression factor of the gun, which is the ratio of the cathode area to the cross sectional area of the final electron beam. Typical compression factors are 5-20.
Electron beams used for linear RF devices typically employ one of two types of magnetic focusing, which act in addition to the initial electrostatic focusing of a Pierce electron gun, whereby a stream of emitted electrons is initially focused to a region of minimum beam diameter. The first type of magnetic focusing is Brillouin focusing, where the magnitude of the magnetic field in the circuit section of the device precisely balances the space charge repulsion forces within the static beam. An embodiment of such a device is shown in FIG. 1. Electrostatic focusing is used to guide the electron beam from the cathode emitting surface to a point within the anode beam tunnel. A minimum diameter is achieved, and if a counteracting magnetic field were not applied, the beam would begin to diverge due to space charge forces. In Brillouin magnetically focused devices, an axial magnetic field is imposed at the location of the minimum diameter that balances the space charge forces and facilitates transport of the beam through the device.
Unfortunately, the balance between the space charge force tending to expand the beam and the magnetic force tending to confine the beam is no longer equal when electrostatic bunching of electrons occurs, as is required to transform beam power into RF power. Consequently, the beam will expand in regions of high electron density, eventually resulting in impact of electrons with the walls of the beam tunnel. This can result in destruction of the device unless the power deposited is limited. Therefore, Brillouin focused devices are limited in the average RF power and pulse lengths that can be generated.
The alternative is to use convergent, confined flow focusing, as shown in FIG. 2. With confined flow focusing, the magnetic field encompasses the cathode regions of the device where the electron beam is generated. A combination of magnetic and electrostatic focusing is used to guide the electron beam from the cathode into the beam tunnel. With confined flow focusing, the magnetic field can be higher than is required for balancing the space charge forces in the static beam. In typical devices, the magnetic field is 2-3 times the Brillouin value. With confined flow focusing, the convergent electron beam will be contained as it traverses the beam tunnel, even in the presence of electron bunching as used to generate RF power. Consequently, confined flow focused devices are capable of high average power operation.
In typical single beam devices, the magnetic field is generated from a solenoid or permanent magnet symmetrically located with respect to the electron beam, which produces a magnetic field that is radially symmetric about the electron beam, which is typically located on the main axis of the device. This radially symmetric field is necessary for the electron beam to follow its non-divergent axial path. The magnitude and shape of the field in the cathode-anode region is controlled using an iron enclosure around the main solenoid or permanent magnet with an aperture through end plates perpendicular to the device axis, allowing field penetration into the cathode-anode region. Auxiliary coils or permanent magnets may also be used in the cathode-anode region to control the shape and magnitude of the field.
While this works well for single beam devices having a beam tunnel symmetrically located with respect to the magnetic field axis, problems occur for electron guns where the cathode-anode region is radially displaced from the device axis. A radial gradient, or shear, in the magnetic field in the cathode-anode region distorts the magnetic focusing, preventing operation of the device. In order to realize a multiple beam device, it is necessary for most cathode-anode structures to be radially displaced from the device axis.
In light of these limitations, the need for a high power, multiple beam klystrons with confined flow focusing for use with high frequency RF sources is clear.